基于贝叶斯决策模型的热轧卷筒电机故障诊断  

Fault diagnosis of reel motor in hot continuous rolling based on Bayesian decision

作　　者：谢丰 付文鹏[2] 李阳 谢向群[2] 李维刚[1] XIE Feng;FU Wen-peng;LI Yang;XIE Xiang-qun;LI Wei-gang(School of Information Science and Engineering.Wuhan University of Science and Technology,Wuhan 430081,Hubei,China;Hot Rolling Plant,Meishan Iron and Steel Co.,Nanjing 210039,Jiangsu,China)

机构地区：[1]武汉科技大学信息科学与工程学院,湖北武汉430081 [2]上海梅山钢铁股份有限公司热轧厂,江苏南京210039

出　　处：《中国冶金》2021年第4期68-73,共6页China Metallurgy

基　　金：国家自然科学基金资助项目(51774219)。

摘　　要：卷筒电机是卷筒的动力来源,对钢卷卷形质量和卷取稳定性有着重要影响。针对热连轧机组卷取工序中卷筒电机的鼠笼条断裂问题,通过从现场采集卷筒电机电流信号,并经过信号截取、滤波、拟合等操作提取出信号特征,利用多元高斯模型进行建模,并基于最小风险的贝叶斯分类算法对其进行故障诊断。试验结果表明,采用最小风险贝叶斯决策模型可以较好地识别卷筒电机的鼠笼条断裂故障,与支持向量机、Fisher等分类器相比,该方法具有较好的分类效果,且通过调节风险系数能够增强对异常样本的捕获能力。The reel motor is the power source of the reel, which has an important influence on the coil shape quality and coiling stability. Aiming at the squirrel cage broken problem of the reel motor in the coiling process of the hot strip mill, the current signal of the reel motor is collected from the steel mill, and the signal characteristics are extracted through signal interception, filtering, fitting and other operations. The multivariate Gaussian model is used for modeling, and the fault diagnosis is performed based on the Bayesian classification algorithm with minimum risk. Experimental results prove that the Bayesian decision based on minimum risk can better diagnose the failure phenomenon of the squirrel cage strip of the reel motor. Compared with support vector machines and Fisher classifiers, the proposed method has a good classification effect and the ability to capture abnormal samples can be enhanced by adjusting the risk coefficient.

关 键 词：卷筒电机 贝叶斯分类器 支持向量机 FISHER分类器 故障诊断 

分 类 号：TG333.24[金属学及工艺—金属压力加工]